Add data tool function and improve response printing

Added `make_data_tool` function which is responsible for generating function metadata for openai tools. Also, replaced ordinary `print` of response from OpenAI with `pretty_print_conversation` in `using_mindsdb_inference_with_text2sql_prompt.py` to enhance visibility of conversation.

Author: dusvyat

examples/using_mindsdb_inference_with_text2sql_prompt.py

@@ -1,5 +1,6 @@
 from openai import OpenAI
-from mindsdb_sdk.utils.openai import extract_sql_query, query_database, chat_completion_request
+from mindsdb_sdk.utils.openai import extract_sql_query, query_database, chat_completion_request, \
+    pretty_print_conversation
 
 import mindsdb_sdk
 import os
@@ -90,4 +91,5 @@ def generate_user_prompt(query: str) -> dict:
                                               tool_choice=None)
 response = chat_completion_gpt.choices[0].message.content
 
-print(response)
+pretty_print_conversation(messages)
+

mindsdb_sdk/utils/openai.py

@@ -109,6 +109,64 @@ def make_mindsdb_tool(schema: dict) -> dict:
     }
 
 
+def make_data_tool(
+    model: str,
+    data_source: str,
+    description: str,
+    connection_args: dict
+):
+    """
+    tool passing connection details for datasource to litellm callback
+
+    :param model: model name for text to sql completion
+    :param data_source: data source name
+    :param description: description of the data source
+    :param connection_args: connection arguments for the data source
+
+    :return: dictionary containing function metadata for openai tools
+    """
+    # Convert the connection_args dictionary to a JSON object
+    connection_args_json = json.dumps(connection_args)
+
+    tool_description = f"""
+Queries the provided data source about user data. When calling this function, ALWAYS use the following arguments:
+- model: {model}
+- connection_args: {connection_args_json}
+- data_source: {data_source}
+- description: {description}
+"""
+
+    return {
+        "type":"function",
+        "function":{
+            "name":"get_mindsdb_text_to_sql_completion",
+            "description":tool_description,
+            "parameters":{
+                "type":"object",
+                "properties":{
+                    "model":{
+                        "type":"string",
+                        "description":"llm model name to use for text to sql completion",
+                    },
+                    "data_source":{
+                        "type":"string",
+                        "description":"Data source name",
+                    },
+                    "connection_args":{
+                        "type":"object",
+                        "description":"Connection arguments for the data source",
+                    },
+                    "description":{
+                        "type":"string",
+                        "description":"Description of the data source",
+                    }
+                },
+                "required": ['data_source', 'connection_args', 'model', 'description']
+            }
+        }
+    }
+
+
 def extract_sql_query(result: str) -> str:
     """
     Extract the SQL query from an openai result string